Discharge coefficient of oblique sharp crested weir for free and submerged flow using trained ANN model

Abstract

In the present study, ANN models have been developed to predict the discharge coefficients of oblique sharp-crested weirs for free and submerged flow cases using Borghei et al.’s experimental data. The discharge coefficients predicted by ANN models are then used to predict the discharges. The results so obtained are compared with the traditional regression model analysis performed by Borghei et al. (2003) in which the prediction error in the discharge was found within the range of ±5%. On the other hand, the developed ANN models predict the discharge coefficients as well as discharges within the error range of ±1%. Furthermore, sensitivity analysis of developed ANN models have been carried out for all the parameters (weir height, oblique weir length, head over weir and downstream head over weir) involved in the study and it was found that the weir length (L) is the most and weir height (P) is the least sensitive input variable to ANN-1 model. In the case of ANN-2 model, weir length (L) is the most and downstream head over weir (Hd) is the least sensitive input variable. Keywords: Artificial neural network, Oblique weir, Discharge coefficient, Free flow, Submerged flow, Discharge measurement